Using Data Science To Revolutionize Geological Logging

The University of Western Australia (UWA) and Rio Tinto Iron Ore (RTIO) have entered into a four-year, $6.1 million research partnership to develop innovative data science solutions (artificial intelligence) for automated geological logging to improve mining practice. The partnership, which follows more than 10 years of collaboration between UWA's data science team and RTIO, will employ five full-time researchers and provide training opportunities for a number of industry-driven PhD programmes. Dr Daniel Wedge, from (CDG) in UWA's School of Geosciences, said UWA's expertise will be resorted to help RTIO's mine geology team tackle the challenge of objective well geological materials. "Until recently, geologist's specialists had to manually interpret and document material found in core samples, a process that was time-consuming and challenging," Dr Wedge said. "Our project can use artificial intelligence: machine learning, pc vision, spacial modelling and improvement techniques to integrate disparate borehole information, together with analysis, imagery, geochemical and natural science informationalong side chemical analysis, imagery, geochemical and earth science info, to."RTIO head Dr. Angus McFarlane said the past partnership between UWA and RTIO has led to the commercialisation of UWA's automated downhole image analysis software and three joint patent applications for RTIO-driven machine learning-based geological modelling.